George Hu, Fireworks AI President on AI Inference and Enterprise

而且你会……

And, you know, you'll...

而且当你作为运营者的时候，你一次只能选一个，所以你真的要好好挑。

And when you're an operator, you only get one at a time, and so you better really pick wisely.

我觉得对我来说最重要的是领导者的素质，尤其是如果你要为……

And I think the most important thing for me is the quality leader, especially if you're gonna work for a...

做我现在这种工作的话。

You know, work with this kind of role that I'm doing.

所以我对林恩非常有信心。

And so I I have, yeah, I have a lot of faith in in Lynn.

她真的是一个我为之骄傲的领导者。

She's she's truly, yeah, a a leader I'm proud to work for.

所以我觉得这是一个很好的起点。

So I think that's, that's a good starting point.

第二点我想说的是，我想待在一个我的工作对公司有价值的地方。

And, you know, the second thing I would say is, like, I wanna be in a place where I wanna be in a place where what I do is a value to the company.

你知道吗？

You know?

而且我觉得我刚来的时候，我确实意识到这里有很大的机会，可以把一些……

And I think when I came in here, one of the things I did realize was that there's a huge opportunity to bring some of the some of the...

不是那些所谓的“剧本”。

You know, not the playbooks.

我讨厌那个词，但就是一些，我觉得，我在其他地方学到的一些东西，一些经验教训。

I hate that word, but, like, just some of the, I think, some of the the things that I learned, lessons learned in other places.

还有，就像我之前提到的，心算或者运营模式，是我早期非常关注的事情之一，就是，好。

And, like, some of the stuff I talked about the mental math or operating model as one of the things I really honed in on early, was, okay.

Fireworks的数学逻辑是什么？

What is the math of Fireworks?

就是说，什么能推动这个业务向前发展？

Like, what is gonna move this business forward?

对业务真正重要的三到四个关键杠杆是什么？

What are the three to four levers that really matter for the business?

那我们怎么让所有人都朝着这些目标努力呢？

And and how do we get everyone aligned to make those things work?

所以尤其是，我觉得甚至……

So, you know, especially, I would say even...

对我的领导团队来说，我们花了很多时间思考这个问题，我经常回到我们的核心逻辑上。

For my leadership team, we spend a lot of time thinking about that, and I'm always constantly coming back to, here's our mental math.

这就是我们看待这个业务的方式。

Here's how we think about this business.

所以我们明白什么重要，然后专注做那些事，而不是其他事，因为就像你说的，现在有无数的事情可以去追，从业务优先级的角度来看。

So, like, we understand what's important, and we chase those things and not, you know, other things because there's a gazillion, to your point, there's a gazillion things we could possibly be chasing right now in terms of in terms of the the the priorities of of the business.

从品牌的角度，或者任何公司的品牌来看，我觉得……

In terms of our brand or any any company's brand, I think that...

你看，我觉得有一种……

Look, I think that there is kind of like this...

有一种……

There's a kind of...

关于像活动这样的东西，比如你提到的Dreamforce这种展会，有两种思考方式。

There's two ways to think about, like, anything like an event, which you'd mentioned, like, Dreamforce for a show.

而且我……

And I...

说到营销，也许……

When it comes to marketing, maybe...

让我先把高度提升到营销层面，然后再回到展会细节。

Let me elevate up to the level of marketing first, and then we'll come back to this detail of a show.

我一直都是从定位和信息开始。

I've always started with the positioning and the message.

对吧？

Okay?

我一直是这样开始的，因为我觉得这是……

I've always started with that because I feel like it's...

这是……

You know, it's...

很多时候，人们很容易纯粹从分析的角度来看待营销。

Oftentimes, it's easy to think about marketing purely analytically.

对。

Okay.

比如，我要花一定数量的钱。

Like, I'm gonna spend a certain amount of money.

我要回答关于管道、需求生成、潜在客户或其他什么的问题。

I'm gonna answer amount of pipeline or demand gen or leads or whatever it is.

但我认为正确的起点是定位，而且我觉得把这一点做好非常重要。

But I feel that the right place to start is with positioning, and I think it's really important to to get that right.

而且我觉得如果你把这一点做好了，那就会……

And I think if you get that right, that is something...

因为品牌是每天建立持久、重复价值的东西，无论你在做什么，无论你在做什么。

Because brand is what builds enduring recurring value every day, whatever you're doing on everything.

比如，每天你的品牌都在那里，并且在买家的心中占据正确的位置，你就是在让一切变得更容易。

Like, every day that your brand is out there and it's the right place in the buyer's mind, you're just making everything easier.

就是说，如果你有一个强大的品牌，一切都会变得容易。

Like, it it unlocks everything if you have a great brand.

如果品牌不好，那所有东西都会变得更贵。

If you have a bad brand, then everything gets more expensive.

每一个需求生成都会变得更贵。

Every demand gen becomes more expensive.

所有事情都会变得更难。

Everything gets harder.

所以在Salesforce，我们做了很多...

And so, you know, at Salesforce, we we worked a lot...

我们非常努力地围绕云这个核心身份来定位，然后逐渐成为客户公司或客户平台。

We worked very hard till we got to the place around cloud, position around cloud as kind of a core identity, and then, kind of becoming the customer company or the customer platform.

你知道吗？

You know?

那就是...

And that's...

说实话，这些至今仍然是Salesforce为人所知的核心理念。

These are enduring ideas that Salesforce still is known for today, honestly.

就是……

That...

这些理念已经积累了二十多年的品牌资产。

That's that's, you know, you have twenty plus years of, you know, brand equity built into some of these ideas.

今天这依然非常非常强大。

And that's just super super powerful, you know, today.

你……

You know, you...

或者说，不管他们做多少营销，或者说他们确实做了很多出色的营销。

Or no matter how much marketing they do or don't do and, you know, do some great marketing.

但如果你在考虑建立任何系统来管理你的客户，你一定会去找他们。

But, you know, if you're out there in the world and you're thinking about, you know, creating any system to deploy to manage your customers, you're gonna go to them.

就是……

You know, it's...

不需要...

It doesn't need...

他们...

They...

就是说，他们根本不用多想。

Like, they don't have to think about it.

他们肯定会在你的候选名单上。

They're just gonna be on your shortlist.

他们必须得在你的候选名单上。

They have to be on your shortlist.

这真的是非常非常强大的事情。

That's a really, really, really powerful thing.

而当你往下走，开始按活动层面购买战术性需求时，我觉得你就失去了品牌持久性的问题。

And when you go down the funnel and you start buying tactical demand, you know, on a campaign level, I think you lose the brand enduring brand question.

所以当我创办公司时，我总是先考虑这个问题。

So when I started a company, I always think about that question first.

我去Twilio的时候，根本没担心过需求生成漏斗的事。

When I went to Twilio, you know, I didn't worry about the demand gen funnel.

而且Jeff之前……

And Jeff had...

我从Jeff那里继承了一个很棒的东西，就是他打造的那整套开发者布道势头，这功劳百分之百是Jeff的。

You know, I had inherited something amazing from Jeff, which is the whole developer evangelism momentum that he created, and Jeff gets a 100% credit for that.

我一点功劳都没有。

I don't get any credit for that.

但我确实接手了一个非常好的局面。

But I inherited something really wonderful there.

不过我们当时……

But we were we were...

我们……

We've...

我们自称……

We called ourselves...

当时我们被归类为CPaaS公司。

We were categorized at the time as a CPaaS company.

通信平台即服务。

A communications platform as a service.

这名字听起来就够糟糕的。

It's as bad as it sounds.

就像……

You know, it's like...

并不是说我是一个……

It's it's not like I'm a c...

我……

I...

你好。

Hello.

我是Georgia，我来自一家CPaaS公司。

I'm I'm Georgia, and I'm a part of a CPaaS company.

听起来我好像得了什么病似的。

Sounds like I have a disease, I think, or something.

总之……

Anyway...

CPaaS的意思是，你是在跟电信部门的人打交道。

And, you know, CPaaS meant like you were selling to the telecom person.

就像你说的，我是做通信平台服务的。

Like you said, I do communication platform service.

我会去找负责我们AT&T账单的人谈，随便吧。

I'll go talk to the person that pays our AT and T bill, whatever.

所以我当时特别关注的一件事，就是把公司重新定位为客户互动，客户互动，因为这种说法，当时在Twilio我们觉得特别陌生，但当你……

And so one the things I really focused on was repositioning the company to customer engagements, you know, customer engagement because that language, which we felt really foreign at the time to us at Twilio, but, like, when you...

你会惊讶地发现，当你说同一件事说了一百万遍之后，突然就变得特别自然了。

You'll be surprised when you say something a million times over and over again, suddenly it sounds sounds just sound really natural.

然后现在大家就觉得，当然啦，我们显然就是客户互动平台。

And suddenly now it just says, of course, obvious we're a customer engagement platform.

但那是……

But that was...

这让我们能接触到高层，接触到业务层面，进行更有战略性的对话，让我们能卖出很高的价格。

Got us up to the c level, up to, like, the business level, up to a more strategic conversation, allow us to sell really high.

所以，这比我们做的任何单个活动都重要。

So, like, that that is more important than any individual campaign we did.

然后，当然，我们通过SIGNAL大会放大了这一点，而且有一位非常出色的CMO，萨拉·瓦尔尼，她为我工作，现在在Datadog。

Then, of course, we amplified it with our conference at SIGNAL and had amazing CMO that, worked for me there, Sarah Varni, who's not a Datadog.

她真的是世界级的，我非常感激她在那方面的合作。

She's really world class, you know, really appreciated her as a partner in that.

但我们做到了，不过我们是从定位开始的，就是把定位做对。

But we did that, but we started on the position of, like, getting the position right.

在Fireworks这里，我进来的时候，我觉得网站上的内容……

And here at Fireworks, you know, I walked in and I think the website something...

好像是说“构建、调优、扩展”，而且首页写的是关于极速推理的。

Said something like build tune scale, and it was about blazing fast inference was what the homepage said.

推理确实很重要，但这并不是公司的真正愿景。

And, you know, inference inference is great, but, you know, it's not really the vision of the company actually.

Lynn 有一个愿景，她给它起了个名字，我现在不想说，因为我们用了不同的说法，但真正重要的是把那个说法弄对。

And Lynn had a vision that she called something, I won't say it was now because we use different language, but really getting that language right.

所以现在我们把自己定位成一家专门做智能的公司，因为我们相信未来不只是推理。

And so now we talk of ourselves as a specialized intelligence company because we believe the future is not just about inference.

而是每家公司都用自己的数据来打造自己的模型。

It's about every company taking its data and creating its own model.

这样你就拥有了自己的智慧。

So you own your intelligence.

你不再...

You're not...

你不再是向 Anthropocrypto 租借智慧，而是真正拥有它。

You're renting your intelligence from Anthropocrypto, but now you really own it.

而且它是为你量身定制的，为你优化的。

And it's tailored for you, and it's optimized for you.

这就是……

And that's...

当你能做到这一点的时候，你就可以真正打破成本和质量之间的这种取舍关系。

You know, when you when you can do that, you can really break this kind of trade off you have to have between cost and quality.

因为大多数人会认为，如果你想要质量，那就会非常昂贵。

Because most people, if they think you want quality, it's super expensive.

前沿技术是很贵的。

Frontier is expensive.

但是现在有了专业智能，你可以用一个很棒的开源模型。

But now with specialized intelligence, you can take a great open model.

你可以针对你的使用场景对它进行微调。

You can fine tune it for your use case.

它的质量往往比前沿技术还要好，但价格却是开源模型的水平。

It has, you know, oftentimes better than Frontier quality, but at an open model price point.

这种拥有你自己的智能，而不是租用它的愿景。

And this vision of owning your own intelligence versus renting it.

所以如果你今天看我们的网站，上面写着，拥有你的模型，拥有你的未来。

So if you look at our website today, it says, you know, like, own your model, own your future.

这跟“构建、调优、扩展”相比，是一个非常不同、更感性、更有力的信息。

That's a very different and much more emotional and powerful message than, you know, build, tune, scale.

没什么……

Nothing...

我不是在批评什么，但我们的起点确实很不一样。

And I don't critique anything, but it's like there's a very different place we're starting.

我更多想的是这个，而不是某个需求生成活动的投资回报率。

I think about that a lot more than I think about the ROI of any particular demand gen campaign.

而且我会……

And I would...

而且我在Salesforce当CMO时学到一件事，我记得，比如我的第二次Dreamforce，什么的。

And and, you know, one thing I learned, you know, when I was CMO at Salesforce, I remember, like, my second Dreamforce, whatever.

你知道吗？

You know?

比如，马克把我拉到一边，在展会现场，然后他说……

Like, Mark pulls me aside and milled the show, and he goes...

他指着某个，比如，像灯一样的东西……

He points to some, like, like, light on...

就是墙上那种发光的标志。

Like, whatever, like, lighted sign on the wall.

他说，那上面是什么信息？

He's like, what is the message?

我就说，好吧。

And I was like, okay.

我说了点什么。

I said something.

他说，好吧。

He's like, okay.

那上面写的是什么？

What does that say?

这把椅子传递了什么信息？

What's the message of the chair?

这款鸡尾酒的名字有什么含义？

What's the message of the name of the cocktail?

我们这里用的颜色有什么含义？

What's the message of the, you know, the colors that we use here?

海报上的信息是什么？

What's the message on the poster?

他说，我们做的每件事都必须强化我们想通过这场演出传达的信息。

He's like, everything we do has to reinforce the message that we're trying to get out for this show.

所以我真的把这一点记住了。

And so I've really taken that away.

关键是信息。

It's about the message.

关键是相关性。

It's about relevance.

这关乎你作为一家公司的形象。

It's about who you are as a company.

这关乎争夺人们的心智，而不仅仅是争夺潜在客户。

It's about battling for the mind, not just for, like, not just for the lead.

我在做任何营销时，真的首先会考虑这一点，而且，呃，这个……

And I really, really think about that first and foremost when I do any of my marketing and, and and this...

尤其是在展会上，因为我觉得很容易只关注潜在客户的数量，但这些数字确实必须有。

In a show, especially, because I think it's very easy to get folks in the pipeline numbers, and you have to have them.

这些都很重要。

It's all important.

但更重要的是，离开展会的人对我这家公司会留下什么印象。

But what's more important is what the person that leaves that show thinking about me as a company.

他们会怎么看我？

What do they think of me?

他们会把我当作领导者吗？

Do they think of me as a leader?

他们会怎么看我？

Do they think of me?

他们会把我归到哪个类别？

What category did they put me in?

他们会怎么看待我们公司？

How do they think of me as a company?

而这正是展会的出发点。

And that is that is the starting point for the show.

说实话，我觉得人们对展会有一个被低估的地方，那就是，从某种奇怪的角度来说，你展会的第一受众甚至不是你的客户，而是你的员工，因为那是你培训员工的地方。

And frankly, one of things that people, I think, underappreciate for shows is that in a weird way, like, the number one audience for your show is not even your customer, It's your employee because that's where you train your employees.

而且，每一个来你展会的客户，那个员工……

And for every customer that shows up at your show, the employee that...

那个来你展会的销售，在那一场会面中，一年里还会跟另外50个、100个人谈。

The salesperson that shows up your show is gonna talk to 50 more, a 100 more times people through the course of the year in that one meeting.

如果你能通过那场展会，把信息准确地传达给你的销售，他们就会在接下来的一年里，把那个信息带给每一个客户，从而让那个信息带来100倍的回报。

And if you can get the message right through that show in the mind of your salesperson, they're gonna take that message the rest of the year to every single customer getting a 100 times ROI from that message.

那如果你只是想想看，我其实只关心那少部分真正会来看展会或者甚至在网上观看的客户。

Then if you just think about, I just care about the the the the small percentage of customers that actually ever ever show up for the show or even watch it online.

所以我觉得，确实应该这样去想。

So it's really, I think, thinking about it that way.

我想从你提到的最后一点开始讲，这完全是真的。

I wanna start from like the last point you mentioned, which is like completely true.

因为我在上学的时候，实际上听了一场比尔·马里奥特的演讲，他是万豪酒店的创始人，或者至少是CEO，我不确定是哪一代人创立的。

Because like I, during school, I actually listened to a talk given by like Bill Marriott, who was like the founder of Marriott hotel or like at least the CEO, I'm not sure which generation founded.

对

Felt

然后他基本上是在说，你要照顾好员工，然后员工就会照顾好整个世界。

And like it then basically he was saying that like, you know, you take care of like the staff member and then the staff member take care of the rest of the universe.

所以这就是为什么，我觉得我从来没有真正想过Dreamforce是为了员工，当我路过Croc鞋的时候，那是Dreamforce，那个在谈论Agent的人，比如Agent in the City之类的。

So that's why, like, I feel like I've never really thought about like Dreamforce is like for the employees, you know, when I was walking by the Croc crock shoe, that's a dream force, like the the person who was like talking about the agent, like agent in the city, you know, like agents and so like that.

是的。

Sure.

我确实觉得这有点像公司里每个人的培训项目，让大家知道我们到底在卖什么。

I do I do feel like, you know, it's kind of like a training program for everybody from the company to know what we're actually selling.

而且这真的关乎士气，我觉得特别特别重要。

And then it's like really for the morale, I feel like it's really, really important.

然后大家真的聚在一起学习怎么卖你们的产品，或者公司的核心信息。

And then people really got together to learn how to sell your product or the main message of the company as well.

我之前真没这么想过，直到你提出来，但这确实很重要。

I just never thought about it that way until you point out, but which is very important.

说到几件事，我特别喜欢“掌握自己的智慧”这个理念，我想先从团队产品说起，但我想先讲竞争这块，因为现在基本上，不管是推理还是什么，很多时候，任何技术产品都是……

Speaking of a few things, one thing is I really like the message of owning your intelligence and I wanna start with, have team product and then I wanna start with the competition piece first because of, like, you know, essentially nowadays, like, I feel like whether it's inference or, like, anything, like, a lot of times, like, any technical product is...

一旦发布超过三个月，就会有另一家公司做一模一样的事情，或者竞争一直存在。

Once it's out in the wild for like over three months, there is another company that's gonna be doing exactly the same thing or just like there's a competition going on all the time.

所以对我个人来说，显然我脑子里有两个想法。

So for my personal, obviously, there's two thoughts in my head.

一个是亚马逊的做法，就是专注于客户，忽略竞争，等等等等。

One is like the Amazon approach, which is like, you know, focus on the customer, you know, ignore the competition, yada yada.

那就像是一群人。

That's like a group of people.

第二种情况，就我个人而言，我不怕那些跟我做一模一样事情的人。

The second group of things, like, personal approach is like, I don't fear of the people who are doing the exact same thing with me.

我怕的是那些走完全不同的路线、做完全不同事情的人。

I fear of the people who is taking a completely different route, doing something completely different.

然后如果那样的话，比如说我有两个假设，我选了第一个去追求。

And then if that bad, let's say I have two hypotheses, I pick number one to go after myself.

而第二个就是别人在做完全不同的事情。

And then number two is someone else doing that completely different.

然后我觉得这群人可能是我更害怕的那种。

And then I think that this group of people is probably something that I'm more of like I fear of.

那你呢？

But what about you?

你怎么看待竞争？

How do you think about competition?

你是那种像亚马逊一样的人，还是那种根本不在乎的？

Are you like Amazon person, or are you kind of like just don't care?

或者这里的氛围是什么样的？

Or what's the what's the vibe here?

我职业生涯早期学到的一句话，我一直努力遵循，就是：看到你的竞争对手，忽略你的竞争对手。

You know, the mantra I learned early in my career, which I kind of live by or try to live by, is see your competitor, ignore your competitor.

我觉得这是我职业生涯早期学到的，而且对我来说，这确实是最好的方法。对。

And I think that is, like, what I learned early in my career, and I think that's the best approach for me anyway Mhmm.

也就是说，你必须了解你的竞争对手在做什么。

Which is you have to be aware of your competitors for doing.

是的。

Yes.

你必须知道。

You have to know.

你必须能够在那个领域里竞争，所有这一切。

You have to be able to, like, compete in the field, the whole thing.

但大多数情况下，你最好还是专注于自己的执行、你的公司、你的客户，可能更像亚马逊的做法。

But for the most part, you know, you're gonna be better off thinking about your own execution, your company, who you're doing, your customers, more maybe like the Amazon approach.

所以我不会说，我完全不去想竞争对手。

So I wouldn't say, like, I don't think about my competitors at all.

你得多多少少想一点，但只需要保持警觉就行。

You know, you have to a little bit, but enough to be aware.

但大部分精力，我可以说90%以上，甚至95%，都是放在你自己正在做的事情上，尤其是当你在一个像我所待过的大多数公司这样的地方，我觉得Fireworks也是……

But, like, I would say 90 plus percent, 95% of the energy is really about about, you know, what you are doing, especially when you are in a place like most of the companies I've been, and I think Fireworks is...

Fireworks也是如此，你确实是在努力推广和发展一个市场。

This is true of Fireworks too, which is that you are really trying to evangelize and grow, like, a market.

你明白吗？

You know?

而且我觉得，我们并不是说这里有一个成熟的大市场，我们只是试图分一小杯羹。

And I think that, you know, it's not like we are just here and there's, like, this big established market, and we're just, like, trying to take a small piece of our share of it.

我们更像是……

Like, we are...

我们相信我们正处在开创一个市场的前沿，这个市场是关于专业智能、微调、开放模型的，都是我们常说的那些东西。

We believe we're at the forefront of creating a market for for specialized intelligence, for fine tuning, for, you know, open models, all the things we talk about.

从某种意义上说，我的竞争对手，或者你随便认为的谁……

And in a sense, my competitors or who you ever...

不管你认为我们的竞争对手是谁，他们其实更像是我的盟友，而不是对手，因为我们都在讲同一种语言，而这就是建立市场的方式。

Whoever you think our competitors are, they're actually really more my allies than my competitors because we're all talking about the same language, and and that's how you build a market.

如果你没有……

You know, you don't have a...

如果你没有竞争对手，那这个市场本身也不值得去追求。

If you don't have competitors, you don't have a market worth pursuing anyway.

所以从大局来看，我们都在共同壮大这个市场。

And so in the big picture, like, we're all growing this market together.

我们努力把蛋糕做得越大越好。

We're trying to make the pie as big as possible.

实际上，任何冲突或紧张，那些让人兴奋的事情，或者其他什么，其实都有助于讲述故事，推动人们前进，因为每次我的竞争对手出现在媒体报道里，他们都在谈论这个市场的理念。

And, actually, any, you know, conflict or tension, all those kind of things that gets people excited or whatever it is, that's actually just all helpful to tell the story and get people, moving because every time if one of my competitors is featured in some media piece, they're talking about the message of the market.

每次有人看到这些，他们都在了解开放模型。

And every time someone sees that, they're learning about open models.

他们在了解专业智能。

They're learning about, you know, specialized intelligence.

所以这就是为什么我不会为这些事纠结。

So, you know, that's one of the reasons why I don't get wrapped around the axle around them.

我觉得我们真正想做的更大的事情，不是，你知道的……

I feel like the bigger thing we're trying to do is not, you know...

比如，谁是我长期的竞争对手？

Like, who's my long term competitor?

这是个有趣的问题。

It's an interesting question.

是……

Is it a...

是现在外面的某个推理平台吗？

Is it an inference platform out there today?

不是。

No.

我的长期竞争对手，比如Anthropic和OpenAI，以及所有那些对未来有非常不同愿景的人。

My long term competitor is, like, Anthropic and OpenAI and everyone that has, like, this very different vision of the future.

就像，是不是...

It's like, do...

就是，三家大公司掌控全世界的智力，我们都得依赖他们，向他们租房。

It's like, three companies own the intelligence of the world and we're all gonna be holding to them and rent from them.

就像是《生活多美好》。

Like, you know, it's like like It's a Wonderful Life.

他们就像波特先生那样，一直想让你住在波特斯维尔，住在他们的贫民窟里，向他们租房子。

You know, they're like they're like mister Potter and forever they want you to be in Pottersville in their slum renting the house from them.

你知道的。

You know.

我们则是说，不。

And we were like, no.

其实，你就像乔治·贝利。

Actually, you're like George Bailey.

你真的可以拥有自己的家。

You can actually own your home.

你真的可以拥有自己的智能。

You can actually own your intelligence.

这是不同的信息。

This is a different message.

所以这就是为什么我喜欢说，在未来，我的信念是，每一个Anthropic或API客户最终都会成为Fireworks的客户。

And so, you know, that's why I like to say, you know, in the future, like, my belief is that every one of these anthropic or API customers will eventually be a Fireworks customer.

因为我相信，这就是我们的使命。

Because I I believe that's, like, that's the mission we're on.

所以，让每家公司都能真正拥有自己的智能，拥有自己的差异化优势。

So free every company to really ultimately own their own intelligence, own their differentiation.

过去我们有一些软件，它们并不是特别有差异化，但你的核心业务，你真正做的、真正是你的秘密武器的事情。

And, you know, in the past, you know, we had some software that we used that was, like, you know, not particularly differentiating, but then your core business, the thing that you do that's really your secret sauce.

比如说，如果你是联邦快递或者UPS，你的物流软件，你怎么做这件事，那就是你的差异化优势。

Like, if you're a FedEx or UPS, like, your logistics software, like, how you do that, that's your differentiator.

那就是你的知识产权。

That's that's your IP.

那是你的知识。

That's your knowledge.

它被编码在软件里。

That's encoded in software.

我们现在只是把这个理念转移到智能上。

We're just transferring that idea to intelligence now.

嗯，你必须把你所有的知识产权、所有的数据、所有的差异化优势都编码到你的智能里，因为它每天都要为你的大部分业务做决策。

Well, you have to encode all of your IP, all your data, all your all your differentiation into your intelligence because that is gonna be making decisions every day for large parts of your business.

所以你的知识产权必须在那里。

And so that's where your IP has to be.

我觉得那是……

And I think that's...

这对我来说非常有道理。

Just makes a ton of sense to me.

我觉得这就是我接触的那些企业想要的。

I think that's what businesses want that I talk to.

而且，最终任何在世界上推广这种理念的人，实际上都是我的盟友，而不是我的竞争对手。

And and ultimately, you know, anyone that is promoting that message in the world is ultimately really my ally, actually, not my competitor.

关于知识产权部分，或者关于Anthropic、OpenEye与其他公司的对比，我有个问题想问你。

I have a question for you on the IP part or like on the Anthropic OpenEye versus the rest of the universe part.

这也是我现在在AI圈子里听到的一个观点，就是人们应该拥有自己的模型，或者普遍拥有自己的数据。

This is a message that I'm currently hearing in the AI community as well about like people owning their own like model or owning their own data in general.

当然。

Sure.

我想，你会不会说未来……

I guess, would you say the future...

我觉得这对大公司来说更重要，因为如果你是……

I feel like this matters more towards the bigger companies as, I guess if you are...

我不太确定，也许吧，比如说像那种拥有数据的公司，比如Shopify或者Sephora或者Crocs。

I'm not sure, maybe like, I don't know, like any company that's like have owns the data, let's call it Shopify or whatever, or Sephora or Crocs.

所以我觉得这些公司拥有你的全部数据。

So I feel like those are the company that you own your entire end of the data.

就像，对于像我这样的普通消费者，对吧？

Like, I think like for a regular consumer like myself, right?

我每天都试着用这些。

Like, I mean, I try with try to be the everyday, right?

就是说，它很，我是说，我只关心我自己的模型吗？

Like, it's like, it's very I mean, do I care about only my own model?

是的，我确实有点关心。

Yes, I do, kind of.

但另一方面，我觉得它确实解决了很多问题。

But on the other hand, I feel like it does solve a lot of these problem.

我自己没有能力构建自己的东西。

I don't have the capacity to build my own stuff.

那你怎么看，如果只谈模型本身，大企业和中小企业之间的区别？

So what is your thought on the enterprise versus SMB when it comes to the model only.

我

I

觉得，说实话，我觉得你可以用一个比喻...

think, honestly, I think you could draw a meta...

如果你想要一个简单的类比或比喻，就像网站一样。早期的时候，建网站又难又贵，你需要一个非常技术性的...

If you want just a simple analogy or metaphor, use it like websites, you know, in the sense that, you know, in the early days, like, was kind of hard and expensive, you know, you needed like a very tech...

需要一个专门的团队去建网站。

Like a dedicated team to go build a website.

所以大企业自己建网站，他们打造了很好的体验，彼此之间也有差异化。

And so enterprises built their own website and they build great experiences and they were differentiated from each other.

而且那...

And that's...

你把你很多的分析、洞察、数据，还有整个客户旅程，都建在那个网站里了。

You know, you you built a lot of your your analytics, your your insight, your data, your whole customer journey was built into that property.

然后随着时间推移，技术越来越好，出现了一些工具，比如Wix或者Squarespace，一下子就普及开来了。

And then over time, it gets democratized down because technology gets better and you have whatever it is, Wix or whatever it is, Squarespace, and boom, starts to, you know, get democratized down.

而且随着技术……

And as technology...

随着平台变得更好，提供这种更定制化或半定制化的体验，帮助公司做出特别的东西，变得越来越容易。

As platforms get better, the ability to deliver these types of more bespoke or semi bespoke, you know, experiences that help companies do something very special becomes easier and easier and easier.

在我看来，基础模型，比如那些开源模型和语言模型，越来越好。

And to me, like, the foundational models, like, where the open models, the LMs are getting better and better.

我们在Fireworks这里也在想办法，利用我们的训练平台等技术，降低门槛，让它变得更容易。

And we are figuring out here at Fireworks how to add technology like our training platform to to lower the bar to make it easier and easier.

所以，是的，可能在早期……

And so, yeah, maybe, like, in the very early days...

说实话，甚至在LLM出现之前，谁能做专业化的智能？

Like, honestly, even before LLMs, who could do specialized intelligence?

就是你们提到的那些公司。

It's like companies you mentioned.

他们以前有庞大的机器学习团队，或者自己去构建专门的模型。

They had whole massive ML teams or to go build their own specialized models.

比如，这里的很多人……

Like, a lot of the people here...

Fireworks 的很多创始团队成员都来自 Meta。

A lot of the founding team here at Fireworks came out of Meta.

他们之所以有这个想法，是因为在 Meta 开发了这些基于机器学习的专有算法，然后觉得，哇，这太厉害了。

They got inspired to do this idea because they built these types of ML based proprietary algorithms, you know, in, at Meta, and it's like, oh, this is amazing.

我们获得了巨大的竞争优势。

We are getting this huge competitive advantage.

怎么才能让更多公司也能用上呢？

How can we democratize this to more companies?

后来大语言模型出现了，尤其是开源模型，大家突然意识到，哎呀。

And then LLMs came along and especially open models came along and said, oh, shoot.

这现在就像是我们的 Squarespace、Wix 那样的工具。

This is now kind of like our, you know, sort of like our Squarespace Squarespace WixMillow.

好的。

Okay.

现在有了一个基础，让公司达到这个最终目标变得容易多了。

Now there's a foundation that makes it much, much easier for companies to get to this end goal.

而且我觉得你会看到，随着时间推移，提供自己模型的成本和复杂性会越来越低，直到越来越多的企业拥有它们。

And I think you're just gonna see that the cost over time and the complexity of being able to deliver your own model is lower and lower until more and more business have them.

那这是否意味着我们个人生活中也需要自己的模型呢？

Now does that mean that you and I in our personal life are gonna need our own models?

你知道吗？

You know?

我肯定有些人可能会这么做，但我觉得我们关注的重点绝大多数还是企业。

Like, I mean, I'm sure some people will probably do it, but, like, I think the vast majority of what we're focused on is business.

而且说实话，今天，中型……

And, you know, I would say today, realistically, mid mid...

或者企业级或中型以上的公司，显然，AI原生公司是另一类。

Or enterprise or mid sized companies above, you know, obviously, AI natives are a different category.

但我们今天并不是针对街角的小店。

But, you know, we're not going after today the mom and pop on the street corner.

我觉得离你当地的理发店或披萨店拥有自己的模型还有一段距离。

I think we're a ways away from, you know, your local, you know, barber or pizza shop having their own model.

但我认为要实现我们所说的愿景，这并不必要。

But I don't think that is necessary to achieve the kind of vision we're talking about.

而且在这个领域，变化非常快，格蕾丝。

And, you know, things change a lot in this, you know, in this world, Grace.

谁知道十年后或五年后会是什么样子。

Who knows ten years from now or five years from now what it could look like.

可能比你想的要容易。

It may be easier than you think.

也许如果我是披萨店老板，我只要对我的大语言模型说几句话，砰，我就有了自己的模型。

Maybe if I'm a pizza shop, I just say a bunch of words to my LLM at the moment and boom, I get my get my own model.

所以我知道的一件事就是，这个市场变化非常快。

So, you know, the one thing I know is that things change really fast in this market.

说到这个……

Speaking I...

说到市场变化快，我想聊聊人才的问题，对吧？

Of things trending really fast in the market, like I wanna talk about the talent, right?

还有人才保留，以及整体的人才情况，比如……

And then also the talent retention as well as like just the talent in general, like...

当然。

Sure.

先从组建团队说起吧。我觉得最重要的一点，可能是我格局比较小，对吧？

So starting with ensembling the team, like I think one of the biggest thing, maybe this is me thinking very small, right?

但我觉得，每当你训练一个智能体，它可能会……好吧，我会听它的。

But I feel like whenever you train an agent, the agent will likely, well, I'll listen to it.

但大多数时候，智能体会听你的，它其实没什么判断力。

But most times the agent will listen to you and it doesn't really have a lot of judgment.

它没有感觉之类的。

It doesn't have a feeling yada yada.

另一方面，我觉得我现在被硅谷洗脑了，但我就是喜欢智能体。

And then on the other hand, I feel like I'm so brainwashed by Silicon Valley right now, but I like agents.

但另一方面，还有人的因素，你知道，人是非常复杂的。

And then on the other hand, there's the human component, which is know, human are very complex.

最优秀的员工总是想当CEO，就像最优秀的员工会说，哦，我要出去自己开公司，等等。

People, the best employee always want to be the CEO, Like the best employee are like, oh, I'm going piece outside my own company, yada.

所以这就是关于训练某样东西的问题，你在某种程度上也是在训练一个小竞争对手，因为他们知道公司怎么运作，怎么运转。

And then so this is like the thing about like, you know, when you're training something, you're kind of also training like a mini competitor to a degree, because of like they knows how the company function, how the company work.

然后你看到所有大公司有时候都是从员工离开之前成功的公司，去创建新公司，这个新公司跟原公司做的事情80%相似，但更先进。

And then you see all the biggest companies sometimes is coming out of like, you know, employee left a previous successful company to like create this new company that's kind of like 80% very similar to whatever this company is doing, but it's like more advanced.

那你觉得怎么组建核心团队？

So how do you think about building, assembling the core team?

还有你怎么看待人才培养、人才保留，以及管理这么多人的不同期望，因为每个人的职业路径都不一样。

And then how do you think about talent training as well as like talent retention and also managing so many people's different expectations what they're doing since everybody have a different career path.

作为高管，你怎么保持自己的心理稳定？

And how do you do the mental stability for yourself as executive?

我觉得问题在于……

I feel like the problem of...

其实有两类问题。

It's two kind of problem.

一类是怎么帮别人解决问题。

One kind of problem is how do you solve the problem for other people?

但另一类是你自己的问题，比如你的心态，什么让你夜不能寐，你怎么面对那种一切都很混乱的状态，还得照常陪孩子睡觉之类的。

But the other problem is like your own, like kind of like mental mindset on like, what keeps you at night and how do you really battle the fact that like, everything is chaos at all time that like, you still have to like really go to sleep with your kids, all that kind of stuff.

我的孩子都16、18、20岁了，所以我大概不用哄他们睡觉了。

Well, my kids are 16, 18, 20, so I'm probably not taking my kids to sleep.

现在他们都不是小孩子了。

These days are not little kids anymore.

但他们还是……

But they still...

我开玩笑的。

I'm just joking.

我开玩笑的。

I'm joking.

不是。

No.

你看，我觉得……

Look, I think...

嗯，如果你不介意的话，我想从一个稍微不同的角度开始，然后再回到你的问题，因为我一直在思考这个问题，就是如果AI让我们这么高效，为什么大家反而比以前更忙了？

Well, if you don't mind, I'll start at a slightly different place and then I'll come back to your question because I've been thinking a lot about this I've been thinking a lot about this question of, like, if AI makes us so productive, why is everyone working harder than ever kind of thing?

DevOps悖论嘛。

DevOps paradox thing.

就是整个那一方面的事情。

That whole that whole aspect of things.

而且在一个如果工作……

And how, you know, in a world where if work was...

如果那个……

If the the...

如果说成果，或者说目标……

If the outputs or if the goal...

如果工作真的是有明确边界的，你知道具体要做多少活，那一切就都成立了。

The work was actually just boxed, you knew exactly how much work to do, that would all be true.

但实际上，工作是没有边界的。

Because essentially, work is unboxed.

你想干多少就能干多少，只要你能撑得住。

You can have as much as you want while you can eat.

那么，有了让你更高效的工具，只会让不工作的机会成本变得更高，所以你就比以往任何时候都更觉得有压力，要更努力地工作。

Then having a tool that makes you more productive just makes the opportunity cost of not working higher, and so therefore, just feel more pressure than ever to work harder than ever.

我觉得这大概是我为自己和所有员工思考的最大问题：你怎么决定……

And I think that's probably, you know, the biggest thing that I think I think about for myself and for all of my employees is how do you, you know, how do you decide to...

就是说，在这样一个地方，你总觉得如果你今天不干活，或者这一分钟不干活，别人或者竞争对手就在干活，那你怎么才能有效地经营你的业务呢？

Like, how do you run your business effectively in a place where, you know, this looming sense that if you're not working today or this minute that someone else is or competitor is.

我觉得这确实很难。

And, I think it's tough.

我觉得这是当今最难的事情之一。

I think that's one of the hardest things, out there today.

在……在……

In the, you know, in the...

我在做SaaS的时候，大家好像有一种不成文的默契，就是人们工作到一定程度……

When I was in the SaaS world, there was kind of like this unspoken agreement that people just worked a certain...

他们工作很努力，但大多数时候一周工作五天，偶尔六天，大概五天半吧。

You know, they worked really hard, but they worked mostly five days a week, maybe six here and there, mostly 5.5, let's say.

而且你知道，如果你周末休息一天，也没关系，因为大体上，你的竞争对手可能也在做同样的事情。

And you kinda knew that if you took whatever a day off of the weekend, it was okay because broadly speaking, your competitors were probably, doing the same thing.

显然，现在这已经不是常态了，我们也不可能把精灵放回瓶子里。

And obviously, that's not the norm now, and we're not gonna put the genie back in the bottle.

那你怎么去解决这个问题呢？

And so, you know, how do you how do you how do you, you know, figure that out?

对我而言，我在开头提到过价值观，以及知道自己真正在乎什么。

And, you know, I think for me, you know, I talked about at the beginning about, like, you know, values and kind of knowing, like like, what what matters to you.

你会跟孩子、家人聊这些。

You talk to kids and family and things like that.

所以对我来说，有几件事是真实的。

And so, you know, I think for me, the couple things are true.

第一，尽管我很努力工作，也很在乎项目，想取得成功，但我……

Number one, you know, as much as I work hard and I care about fireworks and I wanna be successful, like like, I...

对我来说，一个非常重要的点是，它不能是……

It's a really important point to me that it can't be...

没有任何工作真的可以占据全部精力。

That no job actually just can be all consuming.

我就是不能……

Like, I just can't be...

我不认为……

I don't think...

我觉得我能成功的一个原因是，正如我所说，我对什么……有清晰的价值观。

I think one reason I've been being able to be successful is because I, as I said, have, like, clear values of what what...

对我来说，就像有一套优先级顺序。

Like, an order of operations for me.

所以呢，我……

And so, you know, like, I...

像家庭这些事，比任何事业都重要，不管付出什么代价。

Like, other things like family sit above, like, whatever career, every career at all cost.

所以……

So...

你明白吗？

You know?

但如果你真要这么做，那你就得承担更大的责任，要把很多事都处理好。

But then if you're gonna do that if you're gonna do that, then it puts a much broader onus on you to get, like, the...

把很多事情都做对。

Get a lot of things right.

你明白我的意思吗？

Does that make sense?

你的思维模式最好要非常准确。

Like, your mental model better be spot on.

比如你的……

Like, your...

你的执行方式最好要……

The way you execute better be...

你最好要非常准确。

You better be spot on.

所以我觉得这就像……

And so I think that that's, like...

对我来说，关键就是确保领导者有更大的责任要更正确。

That's the key for me is making sure that, you know, I think there's a greater there's a greater bonus on the leader to be right more.

我不需要用不同的方式说。

I don't have to say it say it differently.

所以当我在Salesforce的时候，我常常想，嘿。

So, you know, when I was at, Salesforce, I used to think, you know, hey.

如果一个领导能对个百分之七十、百分之七十五，那已经算是很不错的领导了。

If a leader got it right, like, 70% of the time, 75% of the time maybe even, like, were pretty good leader.

你知道，有些人甚至会争辩，比如高管招聘这种事，五五开可能才是常态。

Know, You some people would argue even, like, things like executive hiring, like, fifty fifty is probably the norm.

你知道吗？

You know?

但在这个地方，我觉得你必须比那准得多。

But in this place, you have to be right much more often than that, I think.

所以我觉得实际上是要多花点时间多想一步，这样你就不用事后去追着弥补了。

So I think it's taking actually an extra extra beat to be thoughtful about things so that, you know, you're not, you know, chasing up and trying catching up with it later.

我不知道怎么用更好听的话来说，但我觉得这就是那种情况。

I don't know how to maybe say it in more elegant language than that, but I think that's the dynamic.

而且我不觉得...

And I don't...

我觉得风气是从上面带出来的。

I think, you know, tone comes from the top.

所以如果我不……

So, you know, if I don't...

我还是尽量保持一些不轰炸下属的工作时间，因为我觉得这风气是从上往下传的。

You know, I try to still keep hours where I'm not, like, blasting my people, because I think that just comes from the top.

如果我这么做，他们也会跟着这么做。

If I do it, they're gonna do it down.

所以长远来看，结果如何我们拭目以待。

So, you know, we'll see in the fullness of time how that plays out.

也许以后我会回头看，说，天哪。

Maybe I'll look back and say, oh my god.

我那时候就该一周七天、一天二十四小时地工作，那才是唯一的办法。

I needed to work twenty four seven, seven days a week, and that's just the only way to do it.

但是，但是……

But but...

不。

No.

我觉得这真的只是……

I mean, I think that's that's just honestly...

我想经营的不只是一家成功的公司，而是一家能让社区的每个人、我的客户和员工都能蓬勃发展的公司。

I I I wanna run not just a successful business, but one that is...

就是大家都能一起成长、一起繁荣。

Helps all the people in the community and my customers and employees, like, also flourish.

我觉得如果只是为了赚钱，大家都搞出心理问题什么的，那也不是我想经营的企业。

I think if we just do it and everyone gets, like, mental health issues and all those kind of things, that's not the kind of business I wanna run anyway.

也许该让别的执行官来经营这种公司。

Maybe maybe there's another executive that should run that business.

我也不清楚。

I don't know.

但我不是这么看的。

But that's not how I look at it.

大部分的事情必须要做对。

70% of stuff have to be right.

好的。

Okay.

那你到底是怎么做到的呢？

How do you how do you really get there?

就是靠一堆直觉和反复确认吗？

Is that, like, just, like, a bunch of, like, gut feeling gut checks?

我觉得销售经验越多，就越容易判断对。

I feel like the more sales you are, the more likely you're gonna get stuff right.

因为我觉得有时候这很考验心理素质，你得更懂人，知道他们真实的行为方式，然后还得做很多假设。

Because I think in the other day, it's a lot of psychological stuff that you have to understand people better, like how people will actually behave, and then you have to really assume a lot of stuff.

作为领导者，你是怎么做到70%以上的准确率的？

Like, how do you, like, how do you do, like, the 70% plus, right, as a leader?

对我来说，很多优秀的领导力或者说好的决策，其实就是靠模式匹配，好的模式匹配。

I think for me, I think a lot of I think a lot of good leadership or good decision making comes down to pattern matching, like, good pattern matching.

你现在听起来特别像风险投资人。

You sound very VC right now.

抱歉。

Sorry.

非常什么？

A very what?

风投？

VC?

你现在说话就很像风投。

Is that what you're very VC right now.

哦，抱歉。

Oh, sorry.

哎呀。

Oops.

好吧，也许我错过了我的使命。

Well, maybe I missed my call.

我开玩笑的。

I'm just kidding.

我想我可以回到Benchmark，如果真是这样的话？

I can I can go back to Benchmark, I guess, if that's what?

但是……

But...

不。

No.

我喜欢Benchmark，但我更喜欢做运营。

I love Benchmark, but I like being an operator.

而且我……

And I...

这确实归结到我的模式匹配能力。

It does come down to me to to pattern matching.

说实话，我觉得我特别擅长倾听……

And, I mean, I think honestly, like, I I just feel like I'm I'm attuned to listen well to...

在合适的场合下。

In the right context.

所以我尽量花很多时间跟客户面对面。

So I try and spend a lot of my time in front of customers.

我最近去了纽约，之前去了西雅图。

You know, I was in New York recently, Seattle before that.

我尽量多出差，多跟客户面对面交流。

I try and spend a lot of time on the road in front of customers.

而且我真的努力去听，认真听他们在说什么，他们用的词我也很注意。

And I really try and listen, like, listen well to what they're saying, their words they're using quite a bit.

我觉得我……

And and I feel like I'm...

我职业生涯经验够多了，能听懂他们的话，而且大部分时候能理解对，然后从中获取我需要的洞察，基本上就是做出更好的决策。

I've just had enough career experience to be able to hear what they're saying and interpret it mostly the right way, I guess, you know, and be able to glean the insights I need to from that to basically make better decisions.

我不知道是不是有什么秘诀。

I don't know if there's, like, some secret sauce.

我不是说只有我这样。

I'm not saying I'm unique in this.

我相信有很多高管可能在这方面比我做得更好，但我觉得对我来说，这……

I'm sure there's lot of executives that are probably better than me at this, but I feel like that's, for me, a...

这是我觉得对我很有力量的一件事。

Something that I think is powerful for me.

另外我想说的是，我小时候上的是公立学校。

The other thing I would say is that, you know, when I was when I was a kid, I was in, I was in public school.

我上四年级的时候，他们把我们……

I was in the fourth grade, and they pull us...

他们把我们，大概十个人吧，拉到一个特别的项目里，不管那是什么项目。

They pulled, I don't know, like, 10 of us aside into this special program, whatever this program was.

这个项目基本上就是教我们怎么不受限制地思考。

And this program, they basically just tried to, like, teach us how to think unconstrained.

还挺有意思的。

It was kind of interesting.

然后他们给了我们一些东西。

So they gave us these things.

不知道你知不知道那种叫“有洞的故事”的东西，就是那种你得……

I don't know if you know these things called stories with holes and stuff like this where you had to, like...

他们问你一个问题，然后答案完全出乎意料，除非你换个思路……

They ask you a question, and then the answer would be completely unintuitive unless you moved...

或者除非你把自己加进去的那个心理假设给去掉。

Or unless you removed a mental assumption that you added yourself into the equation.

我当时就觉得，哦，这就像……

I And was like, oh, this is like...

对我来说，即使是在小时候，这也是一个非常有趣的训练。

It was a super interesting training for me just even as a child.

所以我觉得另一件事就是，我倾向于让思维更自由，更少受约束。

And so I feel like that's the other thing is that I tend to try to be more free in my thinking and less, like, constrained.

这就是为什么我不喜欢“剧本”这个词。

That's why I don't like the word playbook.

我真的尽量从第一性原理去思考问题。

I really try to, like, think through things from first principles.

而且，就像，Salesforce 有一个……

And, like like, Salesforce had a...

他们，他们……

They they...

他们用的这个词叫‘初心’，我觉得这个词用得特别好。

The language they used for this was, like, beginner's mind, and I thought that was really good language.

而且，这也是……

And and, you know, that's, like, one of...

顺便说一句，我给别人的建议就是，拼命找一个适合你、符合你性格的文化。

By the way, piece of advice I give people is, like, you know, search really hard for a culture that fits you and who you are.

我在 Salesforce 找到了，因为我就是……

And I found that at Salesforce because, you know, like, I'm just...

就像我跟你说的，我是一个积极主动、喜欢解决问题的人。

As I told you, I'm a proactive and person who likes to solve problems.

但我以前在另一家公司工作，那家公司后来没做成。

But I used to work for a different company, a company that later did not succeed.

那家公司叫Northpoint Communications。

It's something called Northpoint Communications.

我记得有一次，我做了和后来在Salesforce做的完全一样的事。

And I remember once, like, I did the same thing I actually did at Salesforce.

我还是同一个人，一点没变。

I was the exact same person.

我去找我的副总裁，跟他说有十件事我觉得有问题，需要修复，然后我列出了我的解决方案。

I went to my VP and I said, you know, there's these 10 things that I don't think are, like, they're broken, and here's what I would do to to to to fix them.

然后我把这些发给了管理层，一层层往上发。

And, you know, and I sent it to the manager manager chain, whatever.

结果我的副总裁把我拉到一边，说：以后别再这么干了。

And, like, my VP pulled me aside and said, never do that again.

你绝对不能再这么做了。

You should never do that again.

你只是个基层员工。

You're just an entry level employee.

你不应该跟我说这些事情。

You shouldn't say anything to me as about these things.

我当时就想，好吧。

And I was like, okay.

那我只好离开这里，因为这里不适合我。

Well, I just could get out of here because this is not the place for me.

后来我终于找到了符合我性格的文化，我觉得这真的非常非常重要，因为这种文化也会塑造你，影响你的训练方式和思维方式。

And I eventually found a culture that fit who I am, and I think that's really, really, really a powerful thing because that also is the culture that will shape you and shape your training and shape your mindset.

生活中有很多事情都取决于心态。

A lot of life comes down to mindset.

比如说亚马逊，你跟亚马逊的人聊聊就知道了。

And so, you know, Amazon, you talk to Amazon.

每一种文化都会以一定的方式影响我们每个人。

Every culture shapes all of us in a certain way.

也许这并不会改变我们是谁。

You know, it doesn't change who we are maybe.

这有个先天和后天的因素。

There's a nature nurture thing.

我们天生就有一定的性格，但环境确实会塑造我们的思维模式。

We are some nature by nature a certain way, but definitely does shape our mental models.

而这些思维模式对我们有非常非常强大的控制力。

And those mental models have a very, very strong powerful hold on us.

所以我觉得，脑子里有好的思维模式，关于怎么看待生活、怎么看待工作、怎么看待人，这已经赢了一半了。

And so I think, getting good mental models in in your head about how you think about life, how you think about work, how think about people, I think that's half the battle for us.

那怎么才能获得这些好的思维模式呢？

And how do you get to those good mental models?

我觉得这是每个阶段的人都该认真思考的问题。

Think is a really important question that anyone in all stages of our community think about.

你怎么看？

What would you say?

我有点...

I have like...

我脑子里总是同时有两个问题。

I always have two questions at the same time.

不知道是不是因为我有多动症。

I don't know if it's because I have ADHD about it.

好，那我的两个问题就是。

Okay, so my two questions.

第一个是，当你考虑工作能力的时候，我觉得职位越高，就越不需要技术或战术方面的东西。

One is like when you're thinking about like the competency at work, like I think when the more of like a leadership position you are, the less technical it gets or tactical.

我想不是战术，而是技术方面你得懂。

I guess not tactical, more technical you will have to be.

你可能不会亲自去做营销活动。

You probably don't do the marketing campaign.

你可能不会真的去写代码。

You probably don't do these coding, actually coding.

但现在，在AI时代，很多CEO其实都会写代码。

But nowadays, like all the CEO actually code, but like in the AI era.

所以我想聊聊，你觉得在领导层的时候，你竞争的核心能力是什么？

So I want to talk about like, would you say like what is like the right competency that you're competing with when you're at, like, the leadership level?

比如说在C级高管的位置上。

Let's say when you're at a c suite.

那更多是心理上的较量，还是作为公司领导者真正在做什么？

Is that, like, more of, like, the psychological battleground or, like, what is happening as, like, the leader of a company?

我还有第二个问题，但，对，我也不知道。

And I have a second question, but, like, yeah, I don't know.

可能你会忘掉，因为我……

Maybe you will forget because, like, I

我觉得确实有，确实有，确实有各种不同的领导力思维模式在流传。

don't I think there's definitely there's definitely there's definitely different mental models for leadership that are floating around.

你有那种……

You have kind of like...

我们之前聊了很久关于创始人模式，你得亲自下场去做，然后要处于那种……

You know, we were talking for a while about founder mode, and you've gotta get in there and do it yourself and kind of be in the...